基于神经网络的配送路径优化算法

An Algorithm for Distribution Route Optimization Based on Neural Network

将Hopfield神经网络算法的能量函数值做为模拟退火算法的初始值,使用模拟退火算法以一定概率接收较差值的方法把结果反馈给神经网络,从而克服Hopfield网络算法容易陷入局部最优解的缺点。然后把改进后的算法应用到配送路径优化中,通过对比原有算法,最终得到一种快速、高效的启发式新算法。

In this paper, by making the energy function value of the Hopfield neural network algorithm the initial value for the simulated an- nealing algorithm, we corrected the tendency of the former algorithm toward local optimization and then by applying the improved algorithm to a numerical ease of distribution route optimization and by comparing its result with that of the original algorithm, discovered a new and more rapid heuristie algorithm.